
Investors should look to AMD as it utilizes "agentic AI" and open-source software to rapidly close the competitive gap with NVIDIA, making it a high-conviction play for automated performance optimization. While Meta Platforms (META) faces geopolitical headwinds in AI acquisitions, its aggressive pivot toward space-based solar and long-duration battery storage positions the company as a leader in solving AI's massive energy constraints. In the private markets, Icon is a prime target for disruption in the construction sector, having secured a $201 million military contract and achieving costs below the national average for 3D-printed housing. Rainmaker offers a unique climate-tech opportunity by providing weather modification services to the ski industry and governments, with plans to expand into desalination by 2027. For real estate exposure, focus on "experience-based" retail in business-friendly hubs like Nashville, Austin, and Florida, where safety and high barriers to entry are driving significant rental premiums.
• Acquisition Blocked: China has ordered Meta to unwind its $2 billion acquisition of the AI platform Manus. Despite Manus relocating to Singapore, Beijing cited violations of investment rules. • Leverage Points: The Manus team includes members born and raised in China with family still in the country, which was noted as a potential leverage point for the CCP. • Space Solar Project: Meta is partnering with Overview Energy to beam up to one gigawatt of solar power from orbit to Earth to provide 24/7 power for data centers. • Energy Storage: Partnering with Noon Energy to deploy one gigawatt of ultra-long duration storage batteries.
• Geopolitical Risk: The forced unwinding of the Manus deal highlights the "fiat" regulatory risk of acquiring teams with Chinese origins, even if they have relocated. • Infrastructure Innovation: Meta is aggressively pursuing non-traditional energy sources (space solar) to solve the power constraints of massive AI data centers.
• Agentic AI in Software: AMD is using "agentic loops" to automate performance optimizations and bug fixes. This allows them to optimize customer models 24/7 without manual intervention. • Open Source Advantage: AMD’s ethos of open-source code allows frontier AI models to "see" their source code, making it easier for AI to write specs and code for AMD hardware compared to closed ecosystems. • Heterogeneous Computing: New chips (like the Strix Halo) integrate CPU, GPU, and NPU. AMD is focused on a "seamless" runtime that bounces workloads between these components based on efficiency. • Forward Deployed Engineering (FDE): AMD has hundreds of engineers working directly with customers to execute software development from the "customer backwards."
• Software Parity: AMD is using AI to close the software gap with competitors (like NVIDIA/CUDA) faster than previously possible through automated optimization. • Edge AI: The focus on NPUs in laptops suggests a strong push for local, real-time AI processing (transcription, translation) rather than relying solely on the cloud.
• 3D Printed Construction: The company has built over 250 structures, including a 100-home neighborhood. • Military Contracts: Recently secured a $201 million contract for barracks. They are currently printing 10 barracks in six months at Fort Bliss, significantly faster than traditional construction. • Cost Efficiency: Icon claims their wall systems are now priced below the national average for traditional construction. • Interplanetary Goals: Working on the first structure for the moon, utilizing "in-situ" resource utilization (using moon dust/dirt as building material).
• Construction Disruption: 3D printing is moving from "niche/unique" designs to a scalable solution for the national housing deficit and rapid military deployment. • Vertical Integration: Unlike failed predecessors, Icon’s success is attributed to owning the design, the robots, and the material science (concrete/mortars).
• Weather Modification: The company uses drones to seed clouds with silver iodide (and a new naturally occurring protein agent) to induce snow and rain. • Precision Targeting: They have reduced the targeting area from hundreds of square miles to approximately 8,000 acres, with a goal of hitting 1,000-acre targets (ideal for ski resorts). • Business Model: Uses "value-based pricing" for B2B (charging per inch of snow/gallon of water) and fixed fees for government contracts. • Future Expansion: Planning to enter the Desalination (DeSal) market by Q1 2027 and automated soil tilling by 2028.
• Climate Tech Opportunity: As water scarcity increases, weather modification is becoming a legitimate infrastructure service for state governments and the "Alpine" (ski) industry. • Regulatory Hurdles: While popular in the West (Utah/California), the company faces "chemtrail" conspiracy pushback in the Eastern U.S.
• Real Estate Strategy: Focuses on "high barrier to entry" retail and hospitality (e.g., The Grove, Miramar). • Self-Funded Model: Operates primarily on its own balance sheet rather than raising outside LP funds, allowing for faster decision-making and "rule-breaking" designs (like the Miramar train). • Geographic Shift: Moving capital away from Los Angeles City due to over-regulation and safety concerns, favoring Nashville, Austin, and Florida, as well as business-friendly CA enclaves like Glendale and Calabasas.
• Retail Resilience: Despite the "e-commerce kills retail" narrative, high-end, "experience-focused" physical locations continue to see strong growth (18% CAGR mentioned). • Safety as a Metric: Safety and cleanliness are now primary drivers for commercial real estate value; properties in "safe" zones (Century City) command significantly higher rents than those in neglected urban centers.
• The "Gundo" Defense/Hard Tech Hub: El Segundo and the surrounding "Greater Gundo" area continue to be the center for "sophomore" hard-tech companies (Anduril, Varda, Rainmaker, etc.). • AI Inference vs. Weights: There is a shifting strategic view that inference capacity (the ability to run models) is becoming a more important competitive advantage than the model weights themselves. • Placebo Productivity: Discussion of a study suggesting that "assigned sleep quality" (believing you slept well) significantly predicts cognitive performance, regardless of actual sleep—highlighting the value of a "high-performance mindset" in business.

By John Coogan & Jordi Hays
Technology's daily show (formerly the Technology Brothers Podcast). Streaming live on X and YouTube from 11 - 2 PM PST Monday - Friday. Available on X, Apple, Spotify, and YouTube.